generalarmoryfandomcom-20200214-history
User blog:James doyle/In the beginning
Well, it's under way - my heraldry wiki, a project to transcribe the whole of Burke's General Armory, is now on it's fourth day of existence. Actually of course, I've been working at it for quite a lot longer than four days. And this is going to take a long time - a really long time - a really, really long time - to complete. A few months ago I discovered that the general Armory, originally published in the 19th century, and still in print (albeit it would cost around £100 to buy all the volumes) was available free and online through the Open Archive project. I could download a PDF file scanned from the original publication, and also a text file created by running the same through an OCR programme. That gave me the idea - I could actually use that information to createn a massive online ordinary - a reverse key if you like, to allow you to take a picture of a shield, and identify whose arms it is from the colours and charges used. I could also do some statistical analysis and research: heraldic dictionaries are crammed with terms, some of which are familair and frequent, others of which are rarer. By breaking down the information from the General Armory, I could find out which terms really are used, and which just inventions of the heralds; are crosses more popular than lions as charges, and is it really true, as it appears to be that rampant lions are far more popular than any other? Which colours are the most popular? And which the least? The first step was to begin proofreading the OCR-generated text; as anyone who has worked with OCR files will know, they are prone to misunderstand words, especially when the scan is full of scanning flaws (GA? Check), on old, yellowing paper (GA? Check), and full of obscure words (GA? Check). It took a couple of weeks to do this proofreading and correction just for the section of names beginning with A, about 36 pages out of 1,300. This resulted in about 2,400 entries; which leads me to believe there are about 70-80,000 entries altogether. The next stage was to tabulate the entries into a spreadsheet. Then the arms had to be broken down by field and charge, correcting spellings and ensuring a consistent style as I went (this is an ongoing process, as I learn more about heraldic synonyms, and also Sir Bernard Burke's preferences and whims). All this completed, I began the wiki, and started transferring the information from the spreadsheet to the wiki. So, although as of writing there are just over 100 pages on the wiki, I already have the data for all of the A names ready to go - all 2,435 entries! So, this is just the beginning. In a month or two, I'll have finished transferring A to the wiki. And then I can begin proofreading and tabulating B - which is about three times as long... Category:Blog posts